TFDS اکنون از فرمت Croissant 🥐 پشتیبانی می کند! برای دانستن بیشتر مستندات را بخوانید.

این صفحه به‌وسیله ‏Cloud Translation API‏ ترجمه شده است.

تاپاکو

منابع:

تمام زبانها

برای بارگذاری این مجموعه داده در TFDS از دستور زیر استفاده کنید:

ds = tfds.load('huggingface:tapaco/all_languages')

شرح :

A freely available paraphrase corpus for 73 languages extracted from the Tatoeba database. Tatoeba is a
crowdsourcing project mainly geared towards language learners. Its aim is to provide example sentences and translations for particular
linguistic constructions and words. The paraphrase corpus is created by populating a graph with Tatoeba sentences and equivalence links
between sentences “meaning the same thing”. This graph is then traversed to extract sets of paraphrases. Several language-independent
filters and pruning steps are applied to remove uninteresting sentences. A manual evaluation performed on three languages shows
that between half and three quarters of inferred paraphrases are correct and that most remaining ones are either correct but trivial, or
near-paraphrases that neutralize a morphological distinction. The corpus contains a total of 1.9 million sentences, with 200 – 250 000
sentences per language. It covers a range of languages for which, to our knowledge, no other paraphrase dataset exists.

مجوز : Creative Commons Attribution 2.0 Generic
نسخه : 1.0.0
تقسیم ها :

شکاف	مثال ها
`'train'`	1926192

امکانات :

{
    "paraphrase_set_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "paraphrase": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lists": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "tags": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "language": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

af

برای بارگذاری این مجموعه داده در TFDS از دستور زیر استفاده کنید:

ds = tfds.load('huggingface:tapaco/af')

شرح :

A freely available paraphrase corpus for 73 languages extracted from the Tatoeba database. Tatoeba is a
crowdsourcing project mainly geared towards language learners. Its aim is to provide example sentences and translations for particular
linguistic constructions and words. The paraphrase corpus is created by populating a graph with Tatoeba sentences and equivalence links
between sentences “meaning the same thing”. This graph is then traversed to extract sets of paraphrases. Several language-independent
filters and pruning steps are applied to remove uninteresting sentences. A manual evaluation performed on three languages shows
that between half and three quarters of inferred paraphrases are correct and that most remaining ones are either correct but trivial, or
near-paraphrases that neutralize a morphological distinction. The corpus contains a total of 1.9 million sentences, with 200 – 250 000
sentences per language. It covers a range of languages for which, to our knowledge, no other paraphrase dataset exists.

مجوز : Creative Commons Attribution 2.0 Generic
نسخه : 1.0.0
تقسیم ها :

شکاف	مثال ها
`'train'`	307

امکانات :

{
    "paraphrase_set_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "paraphrase": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lists": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "tags": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "language": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

ar

برای بارگذاری این مجموعه داده در TFDS از دستور زیر استفاده کنید:

ds = tfds.load('huggingface:tapaco/ar')

شرح :

A freely available paraphrase corpus for 73 languages extracted from the Tatoeba database. Tatoeba is a
crowdsourcing project mainly geared towards language learners. Its aim is to provide example sentences and translations for particular
linguistic constructions and words. The paraphrase corpus is created by populating a graph with Tatoeba sentences and equivalence links
between sentences “meaning the same thing”. This graph is then traversed to extract sets of paraphrases. Several language-independent
filters and pruning steps are applied to remove uninteresting sentences. A manual evaluation performed on three languages shows
that between half and three quarters of inferred paraphrases are correct and that most remaining ones are either correct but trivial, or
near-paraphrases that neutralize a morphological distinction. The corpus contains a total of 1.9 million sentences, with 200 – 250 000
sentences per language. It covers a range of languages for which, to our knowledge, no other paraphrase dataset exists.

مجوز : Creative Commons Attribution 2.0 Generic
نسخه : 1.0.0
تقسیم ها :

شکاف	مثال ها
`'train'`	6446

امکانات :

{
    "paraphrase_set_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "paraphrase": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lists": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "tags": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "language": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

az

برای بارگذاری این مجموعه داده در TFDS از دستور زیر استفاده کنید:

ds = tfds.load('huggingface:tapaco/az')

شرح :

A freely available paraphrase corpus for 73 languages extracted from the Tatoeba database. Tatoeba is a
crowdsourcing project mainly geared towards language learners. Its aim is to provide example sentences and translations for particular
linguistic constructions and words. The paraphrase corpus is created by populating a graph with Tatoeba sentences and equivalence links
between sentences “meaning the same thing”. This graph is then traversed to extract sets of paraphrases. Several language-independent
filters and pruning steps are applied to remove uninteresting sentences. A manual evaluation performed on three languages shows
that between half and three quarters of inferred paraphrases are correct and that most remaining ones are either correct but trivial, or
near-paraphrases that neutralize a morphological distinction. The corpus contains a total of 1.9 million sentences, with 200 – 250 000
sentences per language. It covers a range of languages for which, to our knowledge, no other paraphrase dataset exists.

مجوز : Creative Commons Attribution 2.0 Generic
نسخه : 1.0.0
تقسیم ها :

شکاف	مثال ها
`'train'`	624

امکانات :

{
    "paraphrase_set_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "paraphrase": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lists": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "tags": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "language": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

بودن

برای بارگذاری این مجموعه داده در TFDS از دستور زیر استفاده کنید:

ds = tfds.load('huggingface:tapaco/be')

شرح :

A freely available paraphrase corpus for 73 languages extracted from the Tatoeba database. Tatoeba is a
crowdsourcing project mainly geared towards language learners. Its aim is to provide example sentences and translations for particular
linguistic constructions and words. The paraphrase corpus is created by populating a graph with Tatoeba sentences and equivalence links
between sentences “meaning the same thing”. This graph is then traversed to extract sets of paraphrases. Several language-independent
filters and pruning steps are applied to remove uninteresting sentences. A manual evaluation performed on three languages shows
that between half and three quarters of inferred paraphrases are correct and that most remaining ones are either correct but trivial, or
near-paraphrases that neutralize a morphological distinction. The corpus contains a total of 1.9 million sentences, with 200 – 250 000
sentences per language. It covers a range of languages for which, to our knowledge, no other paraphrase dataset exists.

مجوز : Creative Commons Attribution 2.0 Generic
نسخه : 1.0.0
تقسیم ها :

شکاف	مثال ها
`'train'`	1512

امکانات :

{
    "paraphrase_set_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "paraphrase": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lists": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "tags": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "language": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

بر

برای بارگذاری این مجموعه داده در TFDS از دستور زیر استفاده کنید:

ds = tfds.load('huggingface:tapaco/ber')

شرح :

A freely available paraphrase corpus for 73 languages extracted from the Tatoeba database. Tatoeba is a
crowdsourcing project mainly geared towards language learners. Its aim is to provide example sentences and translations for particular
linguistic constructions and words. The paraphrase corpus is created by populating a graph with Tatoeba sentences and equivalence links
between sentences “meaning the same thing”. This graph is then traversed to extract sets of paraphrases. Several language-independent
filters and pruning steps are applied to remove uninteresting sentences. A manual evaluation performed on three languages shows
that between half and three quarters of inferred paraphrases are correct and that most remaining ones are either correct but trivial, or
near-paraphrases that neutralize a morphological distinction. The corpus contains a total of 1.9 million sentences, with 200 – 250 000
sentences per language. It covers a range of languages for which, to our knowledge, no other paraphrase dataset exists.

مجوز : Creative Commons Attribution 2.0 Generic
نسخه : 1.0.0
تقسیم ها :

شکاف	مثال ها
`'train'`	67484

امکانات :

{
    "paraphrase_set_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "paraphrase": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lists": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "tags": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "language": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

bg

برای بارگذاری این مجموعه داده در TFDS از دستور زیر استفاده کنید:

ds = tfds.load('huggingface:tapaco/bg')

شرح :

A freely available paraphrase corpus for 73 languages extracted from the Tatoeba database. Tatoeba is a
crowdsourcing project mainly geared towards language learners. Its aim is to provide example sentences and translations for particular
linguistic constructions and words. The paraphrase corpus is created by populating a graph with Tatoeba sentences and equivalence links
between sentences “meaning the same thing”. This graph is then traversed to extract sets of paraphrases. Several language-independent
filters and pruning steps are applied to remove uninteresting sentences. A manual evaluation performed on three languages shows
that between half and three quarters of inferred paraphrases are correct and that most remaining ones are either correct but trivial, or
near-paraphrases that neutralize a morphological distinction. The corpus contains a total of 1.9 million sentences, with 200 – 250 000
sentences per language. It covers a range of languages for which, to our knowledge, no other paraphrase dataset exists.

مجوز : Creative Commons Attribution 2.0 Generic
نسخه : 1.0.0
تقسیم ها :

شکاف	مثال ها
`'train'`	6324

امکانات :

{
    "paraphrase_set_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "paraphrase": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lists": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "tags": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "language": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

bn

برای بارگذاری این مجموعه داده در TFDS از دستور زیر استفاده کنید:

ds = tfds.load('huggingface:tapaco/bn')

شرح :

A freely available paraphrase corpus for 73 languages extracted from the Tatoeba database. Tatoeba is a
crowdsourcing project mainly geared towards language learners. Its aim is to provide example sentences and translations for particular
linguistic constructions and words. The paraphrase corpus is created by populating a graph with Tatoeba sentences and equivalence links
between sentences “meaning the same thing”. This graph is then traversed to extract sets of paraphrases. Several language-independent
filters and pruning steps are applied to remove uninteresting sentences. A manual evaluation performed on three languages shows
that between half and three quarters of inferred paraphrases are correct and that most remaining ones are either correct but trivial, or
near-paraphrases that neutralize a morphological distinction. The corpus contains a total of 1.9 million sentences, with 200 – 250 000
sentences per language. It covers a range of languages for which, to our knowledge, no other paraphrase dataset exists.

مجوز : Creative Commons Attribution 2.0 Generic
نسخه : 1.0.0
تقسیم ها :

شکاف	مثال ها
`'train'`	1440

امکانات :

{
    "paraphrase_set_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "paraphrase": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lists": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "tags": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "language": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

br

برای بارگذاری این مجموعه داده در TFDS از دستور زیر استفاده کنید:

ds = tfds.load('huggingface:tapaco/br')

شرح :

A freely available paraphrase corpus for 73 languages extracted from the Tatoeba database. Tatoeba is a
crowdsourcing project mainly geared towards language learners. Its aim is to provide example sentences and translations for particular
linguistic constructions and words. The paraphrase corpus is created by populating a graph with Tatoeba sentences and equivalence links
between sentences “meaning the same thing”. This graph is then traversed to extract sets of paraphrases. Several language-independent
filters and pruning steps are applied to remove uninteresting sentences. A manual evaluation performed on three languages shows
that between half and three quarters of inferred paraphrases are correct and that most remaining ones are either correct but trivial, or
near-paraphrases that neutralize a morphological distinction. The corpus contains a total of 1.9 million sentences, with 200 – 250 000
sentences per language. It covers a range of languages for which, to our knowledge, no other paraphrase dataset exists.

مجوز : Creative Commons Attribution 2.0 Generic
نسخه : 1.0.0
تقسیم ها :

شکاف	مثال ها
`'train'`	2536

امکانات :

{
    "paraphrase_set_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "paraphrase": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lists": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "tags": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "language": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

حدود

برای بارگذاری این مجموعه داده در TFDS از دستور زیر استفاده کنید:

ds = tfds.load('huggingface:tapaco/ca')

شرح :

A freely available paraphrase corpus for 73 languages extracted from the Tatoeba database. Tatoeba is a
crowdsourcing project mainly geared towards language learners. Its aim is to provide example sentences and translations for particular
linguistic constructions and words. The paraphrase corpus is created by populating a graph with Tatoeba sentences and equivalence links
between sentences “meaning the same thing”. This graph is then traversed to extract sets of paraphrases. Several language-independent
filters and pruning steps are applied to remove uninteresting sentences. A manual evaluation performed on three languages shows
that between half and three quarters of inferred paraphrases are correct and that most remaining ones are either correct but trivial, or
near-paraphrases that neutralize a morphological distinction. The corpus contains a total of 1.9 million sentences, with 200 – 250 000
sentences per language. It covers a range of languages for which, to our knowledge, no other paraphrase dataset exists.

مجوز : Creative Commons Attribution 2.0 Generic
نسخه : 1.0.0
تقسیم ها :

شکاف	مثال ها
`'train'`	518

امکانات :

{
    "paraphrase_set_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "paraphrase": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lists": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "tags": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "language": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

cbk

برای بارگذاری این مجموعه داده در TFDS از دستور زیر استفاده کنید:

ds = tfds.load('huggingface:tapaco/cbk')

شرح :

A freely available paraphrase corpus for 73 languages extracted from the Tatoeba database. Tatoeba is a
crowdsourcing project mainly geared towards language learners. Its aim is to provide example sentences and translations for particular
linguistic constructions and words. The paraphrase corpus is created by populating a graph with Tatoeba sentences and equivalence links
between sentences “meaning the same thing”. This graph is then traversed to extract sets of paraphrases. Several language-independent
filters and pruning steps are applied to remove uninteresting sentences. A manual evaluation performed on three languages shows
that between half and three quarters of inferred paraphrases are correct and that most remaining ones are either correct but trivial, or
near-paraphrases that neutralize a morphological distinction. The corpus contains a total of 1.9 million sentences, with 200 – 250 000
sentences per language. It covers a range of languages for which, to our knowledge, no other paraphrase dataset exists.

مجوز : Creative Commons Attribution 2.0 Generic
نسخه : 1.0.0
تقسیم ها :

شکاف	مثال ها
`'train'`	262

امکانات :

{
    "paraphrase_set_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "paraphrase": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lists": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "tags": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "language": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

cmn

برای بارگذاری این مجموعه داده در TFDS از دستور زیر استفاده کنید:

ds = tfds.load('huggingface:tapaco/cmn')

شرح :

A freely available paraphrase corpus for 73 languages extracted from the Tatoeba database. Tatoeba is a
crowdsourcing project mainly geared towards language learners. Its aim is to provide example sentences and translations for particular
linguistic constructions and words. The paraphrase corpus is created by populating a graph with Tatoeba sentences and equivalence links
between sentences “meaning the same thing”. This graph is then traversed to extract sets of paraphrases. Several language-independent
filters and pruning steps are applied to remove uninteresting sentences. A manual evaluation performed on three languages shows
that between half and three quarters of inferred paraphrases are correct and that most remaining ones are either correct but trivial, or
near-paraphrases that neutralize a morphological distinction. The corpus contains a total of 1.9 million sentences, with 200 – 250 000
sentences per language. It covers a range of languages for which, to our knowledge, no other paraphrase dataset exists.

مجوز : Creative Commons Attribution 2.0 Generic
نسخه : 1.0.0
تقسیم ها :

شکاف	مثال ها
`'train'`	12549

امکانات :

{
    "paraphrase_set_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "paraphrase": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lists": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "tags": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "language": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

cs

برای بارگذاری این مجموعه داده در TFDS از دستور زیر استفاده کنید:

ds = tfds.load('huggingface:tapaco/cs')

شرح :

A freely available paraphrase corpus for 73 languages extracted from the Tatoeba database. Tatoeba is a
crowdsourcing project mainly geared towards language learners. Its aim is to provide example sentences and translations for particular
linguistic constructions and words. The paraphrase corpus is created by populating a graph with Tatoeba sentences and equivalence links
between sentences “meaning the same thing”. This graph is then traversed to extract sets of paraphrases. Several language-independent
filters and pruning steps are applied to remove uninteresting sentences. A manual evaluation performed on three languages shows
that between half and three quarters of inferred paraphrases are correct and that most remaining ones are either correct but trivial, or
near-paraphrases that neutralize a morphological distinction. The corpus contains a total of 1.9 million sentences, with 200 – 250 000
sentences per language. It covers a range of languages for which, to our knowledge, no other paraphrase dataset exists.

مجوز : Creative Commons Attribution 2.0 Generic
نسخه : 1.0.0
تقسیم ها :

شکاف	مثال ها
`'train'`	6659

امکانات :

{
    "paraphrase_set_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "paraphrase": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lists": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "tags": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "language": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

دا

برای بارگذاری این مجموعه داده در TFDS از دستور زیر استفاده کنید:

ds = tfds.load('huggingface:tapaco/da')

شرح :

A freely available paraphrase corpus for 73 languages extracted from the Tatoeba database. Tatoeba is a
crowdsourcing project mainly geared towards language learners. Its aim is to provide example sentences and translations for particular
linguistic constructions and words. The paraphrase corpus is created by populating a graph with Tatoeba sentences and equivalence links
between sentences “meaning the same thing”. This graph is then traversed to extract sets of paraphrases. Several language-independent
filters and pruning steps are applied to remove uninteresting sentences. A manual evaluation performed on three languages shows
that between half and three quarters of inferred paraphrases are correct and that most remaining ones are either correct but trivial, or
near-paraphrases that neutralize a morphological distinction. The corpus contains a total of 1.9 million sentences, with 200 – 250 000
sentences per language. It covers a range of languages for which, to our knowledge, no other paraphrase dataset exists.

مجوز : Creative Commons Attribution 2.0 Generic
نسخه : 1.0.0
تقسیم ها :

شکاف	مثال ها
`'train'`	11220

امکانات :

{
    "paraphrase_set_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "paraphrase": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lists": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "tags": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "language": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

de

برای بارگذاری این مجموعه داده در TFDS از دستور زیر استفاده کنید:

ds = tfds.load('huggingface:tapaco/de')

شرح :

A freely available paraphrase corpus for 73 languages extracted from the Tatoeba database. Tatoeba is a
crowdsourcing project mainly geared towards language learners. Its aim is to provide example sentences and translations for particular
linguistic constructions and words. The paraphrase corpus is created by populating a graph with Tatoeba sentences and equivalence links
between sentences “meaning the same thing”. This graph is then traversed to extract sets of paraphrases. Several language-independent
filters and pruning steps are applied to remove uninteresting sentences. A manual evaluation performed on three languages shows
that between half and three quarters of inferred paraphrases are correct and that most remaining ones are either correct but trivial, or
near-paraphrases that neutralize a morphological distinction. The corpus contains a total of 1.9 million sentences, with 200 – 250 000
sentences per language. It covers a range of languages for which, to our knowledge, no other paraphrase dataset exists.

مجوز : Creative Commons Attribution 2.0 Generic
نسخه : 1.0.0
تقسیم ها :

شکاف	مثال ها
`'train'`	125091

امکانات :

{
    "paraphrase_set_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "paraphrase": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lists": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "tags": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "language": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

el

برای بارگذاری این مجموعه داده در TFDS از دستور زیر استفاده کنید:

ds = tfds.load('huggingface:tapaco/el')

شرح :

A freely available paraphrase corpus for 73 languages extracted from the Tatoeba database. Tatoeba is a
crowdsourcing project mainly geared towards language learners. Its aim is to provide example sentences and translations for particular
linguistic constructions and words. The paraphrase corpus is created by populating a graph with Tatoeba sentences and equivalence links
between sentences “meaning the same thing”. This graph is then traversed to extract sets of paraphrases. Several language-independent
filters and pruning steps are applied to remove uninteresting sentences. A manual evaluation performed on three languages shows
that between half and three quarters of inferred paraphrases are correct and that most remaining ones are either correct but trivial, or
near-paraphrases that neutralize a morphological distinction. The corpus contains a total of 1.9 million sentences, with 200 – 250 000
sentences per language. It covers a range of languages for which, to our knowledge, no other paraphrase dataset exists.

مجوز : Creative Commons Attribution 2.0 Generic
نسخه : 1.0.0
تقسیم ها :

شکاف	مثال ها
`'train'`	10072

امکانات :

{
    "paraphrase_set_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "paraphrase": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lists": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "tags": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "language": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

en

برای بارگذاری این مجموعه داده در TFDS از دستور زیر استفاده کنید:

ds = tfds.load('huggingface:tapaco/en')

شرح :

A freely available paraphrase corpus for 73 languages extracted from the Tatoeba database. Tatoeba is a
crowdsourcing project mainly geared towards language learners. Its aim is to provide example sentences and translations for particular
linguistic constructions and words. The paraphrase corpus is created by populating a graph with Tatoeba sentences and equivalence links
between sentences “meaning the same thing”. This graph is then traversed to extract sets of paraphrases. Several language-independent
filters and pruning steps are applied to remove uninteresting sentences. A manual evaluation performed on three languages shows
that between half and three quarters of inferred paraphrases are correct and that most remaining ones are either correct but trivial, or
near-paraphrases that neutralize a morphological distinction. The corpus contains a total of 1.9 million sentences, with 200 – 250 000
sentences per language. It covers a range of languages for which, to our knowledge, no other paraphrase dataset exists.

مجوز : Creative Commons Attribution 2.0 Generic
نسخه : 1.0.0
تقسیم ها :

شکاف	مثال ها
`'train'`	158053

امکانات :

{
    "paraphrase_set_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "paraphrase": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lists": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "tags": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "language": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

eo

برای بارگذاری این مجموعه داده در TFDS از دستور زیر استفاده کنید:

ds = tfds.load('huggingface:tapaco/eo')

شرح :

A freely available paraphrase corpus for 73 languages extracted from the Tatoeba database. Tatoeba is a
crowdsourcing project mainly geared towards language learners. Its aim is to provide example sentences and translations for particular
linguistic constructions and words. The paraphrase corpus is created by populating a graph with Tatoeba sentences and equivalence links
between sentences “meaning the same thing”. This graph is then traversed to extract sets of paraphrases. Several language-independent
filters and pruning steps are applied to remove uninteresting sentences. A manual evaluation performed on three languages shows
that between half and three quarters of inferred paraphrases are correct and that most remaining ones are either correct but trivial, or
near-paraphrases that neutralize a morphological distinction. The corpus contains a total of 1.9 million sentences, with 200 – 250 000
sentences per language. It covers a range of languages for which, to our knowledge, no other paraphrase dataset exists.

مجوز : Creative Commons Attribution 2.0 Generic
نسخه : 1.0.0
تقسیم ها :

شکاف	مثال ها
`'train'`	207105

امکانات :

{
    "paraphrase_set_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "paraphrase": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lists": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "tags": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "language": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

es

برای بارگذاری این مجموعه داده در TFDS از دستور زیر استفاده کنید:

ds = tfds.load('huggingface:tapaco/es')

شرح :

A freely available paraphrase corpus for 73 languages extracted from the Tatoeba database. Tatoeba is a
crowdsourcing project mainly geared towards language learners. Its aim is to provide example sentences and translations for particular
linguistic constructions and words. The paraphrase corpus is created by populating a graph with Tatoeba sentences and equivalence links
between sentences “meaning the same thing”. This graph is then traversed to extract sets of paraphrases. Several language-independent
filters and pruning steps are applied to remove uninteresting sentences. A manual evaluation performed on three languages shows
that between half and three quarters of inferred paraphrases are correct and that most remaining ones are either correct but trivial, or
near-paraphrases that neutralize a morphological distinction. The corpus contains a total of 1.9 million sentences, with 200 – 250 000
sentences per language. It covers a range of languages for which, to our knowledge, no other paraphrase dataset exists.

مجوز : Creative Commons Attribution 2.0 Generic
نسخه : 1.0.0
تقسیم ها :

شکاف	مثال ها
`'train'`	85064

امکانات :

{
    "paraphrase_set_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "paraphrase": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lists": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "tags": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "language": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

et

برای بارگذاری این مجموعه داده در TFDS از دستور زیر استفاده کنید:

ds = tfds.load('huggingface:tapaco/et')

شرح :

A freely available paraphrase corpus for 73 languages extracted from the Tatoeba database. Tatoeba is a
crowdsourcing project mainly geared towards language learners. Its aim is to provide example sentences and translations for particular
linguistic constructions and words. The paraphrase corpus is created by populating a graph with Tatoeba sentences and equivalence links
between sentences “meaning the same thing”. This graph is then traversed to extract sets of paraphrases. Several language-independent
filters and pruning steps are applied to remove uninteresting sentences. A manual evaluation performed on three languages shows
that between half and three quarters of inferred paraphrases are correct and that most remaining ones are either correct but trivial, or
near-paraphrases that neutralize a morphological distinction. The corpus contains a total of 1.9 million sentences, with 200 – 250 000
sentences per language. It covers a range of languages for which, to our knowledge, no other paraphrase dataset exists.

مجوز : Creative Commons Attribution 2.0 Generic
نسخه : 1.0.0
تقسیم ها :

شکاف	مثال ها
`'train'`	241

امکانات :

{
    "paraphrase_set_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "paraphrase": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lists": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "tags": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "language": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

اتحادیه اروپا

برای بارگذاری این مجموعه داده در TFDS از دستور زیر استفاده کنید:

ds = tfds.load('huggingface:tapaco/eu')

شرح :

A freely available paraphrase corpus for 73 languages extracted from the Tatoeba database. Tatoeba is a
crowdsourcing project mainly geared towards language learners. Its aim is to provide example sentences and translations for particular
linguistic constructions and words. The paraphrase corpus is created by populating a graph with Tatoeba sentences and equivalence links
between sentences “meaning the same thing”. This graph is then traversed to extract sets of paraphrases. Several language-independent
filters and pruning steps are applied to remove uninteresting sentences. A manual evaluation performed on three languages shows
that between half and three quarters of inferred paraphrases are correct and that most remaining ones are either correct but trivial, or
near-paraphrases that neutralize a morphological distinction. The corpus contains a total of 1.9 million sentences, with 200 – 250 000
sentences per language. It covers a range of languages for which, to our knowledge, no other paraphrase dataset exists.

مجوز : Creative Commons Attribution 2.0 Generic
نسخه : 1.0.0
تقسیم ها :

شکاف	مثال ها
`'train'`	573

امکانات :

{
    "paraphrase_set_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "paraphrase": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lists": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "tags": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "language": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

فی

برای بارگذاری این مجموعه داده در TFDS از دستور زیر استفاده کنید:

ds = tfds.load('huggingface:tapaco/fi')

شرح :

A freely available paraphrase corpus for 73 languages extracted from the Tatoeba database. Tatoeba is a
crowdsourcing project mainly geared towards language learners. Its aim is to provide example sentences and translations for particular
linguistic constructions and words. The paraphrase corpus is created by populating a graph with Tatoeba sentences and equivalence links
between sentences “meaning the same thing”. This graph is then traversed to extract sets of paraphrases. Several language-independent
filters and pruning steps are applied to remove uninteresting sentences. A manual evaluation performed on three languages shows
that between half and three quarters of inferred paraphrases are correct and that most remaining ones are either correct but trivial, or
near-paraphrases that neutralize a morphological distinction. The corpus contains a total of 1.9 million sentences, with 200 – 250 000
sentences per language. It covers a range of languages for which, to our knowledge, no other paraphrase dataset exists.

مجوز : Creative Commons Attribution 2.0 Generic
نسخه : 1.0.0
تقسیم ها :

شکاف	مثال ها
`'train'`	31753

امکانات :

{
    "paraphrase_set_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "paraphrase": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lists": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "tags": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "language": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

fr

برای بارگذاری این مجموعه داده در TFDS از دستور زیر استفاده کنید:

ds = tfds.load('huggingface:tapaco/fr')

شرح :

A freely available paraphrase corpus for 73 languages extracted from the Tatoeba database. Tatoeba is a
crowdsourcing project mainly geared towards language learners. Its aim is to provide example sentences and translations for particular
linguistic constructions and words. The paraphrase corpus is created by populating a graph with Tatoeba sentences and equivalence links
between sentences “meaning the same thing”. This graph is then traversed to extract sets of paraphrases. Several language-independent
filters and pruning steps are applied to remove uninteresting sentences. A manual evaluation performed on three languages shows
that between half and three quarters of inferred paraphrases are correct and that most remaining ones are either correct but trivial, or
near-paraphrases that neutralize a morphological distinction. The corpus contains a total of 1.9 million sentences, with 200 – 250 000
sentences per language. It covers a range of languages for which, to our knowledge, no other paraphrase dataset exists.

مجوز : Creative Commons Attribution 2.0 Generic
نسخه : 1.0.0
تقسیم ها :

شکاف	مثال ها
`'train'`	116733

امکانات :

{
    "paraphrase_set_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "paraphrase": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lists": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "tags": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "language": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

gl

برای بارگذاری این مجموعه داده در TFDS از دستور زیر استفاده کنید:

ds = tfds.load('huggingface:tapaco/gl')

شرح :

A freely available paraphrase corpus for 73 languages extracted from the Tatoeba database. Tatoeba is a
crowdsourcing project mainly geared towards language learners. Its aim is to provide example sentences and translations for particular
linguistic constructions and words. The paraphrase corpus is created by populating a graph with Tatoeba sentences and equivalence links
between sentences “meaning the same thing”. This graph is then traversed to extract sets of paraphrases. Several language-independent
filters and pruning steps are applied to remove uninteresting sentences. A manual evaluation performed on three languages shows
that between half and three quarters of inferred paraphrases are correct and that most remaining ones are either correct but trivial, or
near-paraphrases that neutralize a morphological distinction. The corpus contains a total of 1.9 million sentences, with 200 – 250 000
sentences per language. It covers a range of languages for which, to our knowledge, no other paraphrase dataset exists.

مجوز : Creative Commons Attribution 2.0 Generic
نسخه : 1.0.0
تقسیم ها :

شکاف	مثال ها
`'train'`	351

امکانات :

{
    "paraphrase_set_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "paraphrase": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lists": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "tags": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "language": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

می رود

برای بارگذاری این مجموعه داده در TFDS از دستور زیر استفاده کنید:

ds = tfds.load('huggingface:tapaco/gos')

شرح :

A freely available paraphrase corpus for 73 languages extracted from the Tatoeba database. Tatoeba is a
crowdsourcing project mainly geared towards language learners. Its aim is to provide example sentences and translations for particular
linguistic constructions and words. The paraphrase corpus is created by populating a graph with Tatoeba sentences and equivalence links
between sentences “meaning the same thing”. This graph is then traversed to extract sets of paraphrases. Several language-independent
filters and pruning steps are applied to remove uninteresting sentences. A manual evaluation performed on three languages shows
that between half and three quarters of inferred paraphrases are correct and that most remaining ones are either correct but trivial, or
near-paraphrases that neutralize a morphological distinction. The corpus contains a total of 1.9 million sentences, with 200 – 250 000
sentences per language. It covers a range of languages for which, to our knowledge, no other paraphrase dataset exists.

مجوز : Creative Commons Attribution 2.0 Generic
نسخه : 1.0.0
تقسیم ها :

شکاف	مثال ها
`'train'`	279

امکانات :

{
    "paraphrase_set_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "paraphrase": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lists": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "tags": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "language": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

او

برای بارگذاری این مجموعه داده در TFDS از دستور زیر استفاده کنید:

ds = tfds.load('huggingface:tapaco/he')

شرح :

A freely available paraphrase corpus for 73 languages extracted from the Tatoeba database. Tatoeba is a
crowdsourcing project mainly geared towards language learners. Its aim is to provide example sentences and translations for particular
linguistic constructions and words. The paraphrase corpus is created by populating a graph with Tatoeba sentences and equivalence links
between sentences “meaning the same thing”. This graph is then traversed to extract sets of paraphrases. Several language-independent
filters and pruning steps are applied to remove uninteresting sentences. A manual evaluation performed on three languages shows
that between half and three quarters of inferred paraphrases are correct and that most remaining ones are either correct but trivial, or
near-paraphrases that neutralize a morphological distinction. The corpus contains a total of 1.9 million sentences, with 200 – 250 000
sentences per language. It covers a range of languages for which, to our knowledge, no other paraphrase dataset exists.

مجوز : Creative Commons Attribution 2.0 Generic
نسخه : 1.0.0
تقسیم ها :

شکاف	مثال ها
`'train'`	68350

امکانات :

{
    "paraphrase_set_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "paraphrase": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lists": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "tags": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "language": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

سلام

برای بارگذاری این مجموعه داده در TFDS از دستور زیر استفاده کنید:

ds = tfds.load('huggingface:tapaco/hi')

شرح :

A freely available paraphrase corpus for 73 languages extracted from the Tatoeba database. Tatoeba is a
crowdsourcing project mainly geared towards language learners. Its aim is to provide example sentences and translations for particular
linguistic constructions and words. The paraphrase corpus is created by populating a graph with Tatoeba sentences and equivalence links
between sentences “meaning the same thing”. This graph is then traversed to extract sets of paraphrases. Several language-independent
filters and pruning steps are applied to remove uninteresting sentences. A manual evaluation performed on three languages shows
that between half and three quarters of inferred paraphrases are correct and that most remaining ones are either correct but trivial, or
near-paraphrases that neutralize a morphological distinction. The corpus contains a total of 1.9 million sentences, with 200 – 250 000
sentences per language. It covers a range of languages for which, to our knowledge, no other paraphrase dataset exists.

مجوز : Creative Commons Attribution 2.0 Generic
نسخه : 1.0.0
تقسیم ها :

شکاف	مثال ها
`'train'`	1913

امکانات :

{
    "paraphrase_set_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "paraphrase": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lists": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "tags": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "language": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

ساعت

برای بارگذاری این مجموعه داده در TFDS از دستور زیر استفاده کنید:

ds = tfds.load('huggingface:tapaco/hr')

شرح :

A freely available paraphrase corpus for 73 languages extracted from the Tatoeba database. Tatoeba is a
crowdsourcing project mainly geared towards language learners. Its aim is to provide example sentences and translations for particular
linguistic constructions and words. The paraphrase corpus is created by populating a graph with Tatoeba sentences and equivalence links
between sentences “meaning the same thing”. This graph is then traversed to extract sets of paraphrases. Several language-independent
filters and pruning steps are applied to remove uninteresting sentences. A manual evaluation performed on three languages shows
that between half and three quarters of inferred paraphrases are correct and that most remaining ones are either correct but trivial, or
near-paraphrases that neutralize a morphological distinction. The corpus contains a total of 1.9 million sentences, with 200 – 250 000
sentences per language. It covers a range of languages for which, to our knowledge, no other paraphrase dataset exists.

مجوز : Creative Commons Attribution 2.0 Generic
نسخه : 1.0.0
تقسیم ها :

شکاف	مثال ها
`'train'`	505

امکانات :

{
    "paraphrase_set_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "paraphrase": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lists": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "tags": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "language": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

هو

برای بارگذاری این مجموعه داده در TFDS از دستور زیر استفاده کنید:

ds = tfds.load('huggingface:tapaco/hu')

شرح :

A freely available paraphrase corpus for 73 languages extracted from the Tatoeba database. Tatoeba is a
crowdsourcing project mainly geared towards language learners. Its aim is to provide example sentences and translations for particular
linguistic constructions and words. The paraphrase corpus is created by populating a graph with Tatoeba sentences and equivalence links
between sentences “meaning the same thing”. This graph is then traversed to extract sets of paraphrases. Several language-independent
filters and pruning steps are applied to remove uninteresting sentences. A manual evaluation performed on three languages shows
that between half and three quarters of inferred paraphrases are correct and that most remaining ones are either correct but trivial, or
near-paraphrases that neutralize a morphological distinction. The corpus contains a total of 1.9 million sentences, with 200 – 250 000
sentences per language. It covers a range of languages for which, to our knowledge, no other paraphrase dataset exists.

مجوز : Creative Commons Attribution 2.0 Generic
نسخه : 1.0.0
تقسیم ها :

شکاف	مثال ها
`'train'`	67964

امکانات :

{
    "paraphrase_set_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "paraphrase": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lists": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "tags": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "language": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

هی

برای بارگذاری این مجموعه داده در TFDS از دستور زیر استفاده کنید:

ds = tfds.load('huggingface:tapaco/hy')

شرح :

A freely available paraphrase corpus for 73 languages extracted from the Tatoeba database. Tatoeba is a
crowdsourcing project mainly geared towards language learners. Its aim is to provide example sentences and translations for particular
linguistic constructions and words. The paraphrase corpus is created by populating a graph with Tatoeba sentences and equivalence links
between sentences “meaning the same thing”. This graph is then traversed to extract sets of paraphrases. Several language-independent
filters and pruning steps are applied to remove uninteresting sentences. A manual evaluation performed on three languages shows
that between half and three quarters of inferred paraphrases are correct and that most remaining ones are either correct but trivial, or
near-paraphrases that neutralize a morphological distinction. The corpus contains a total of 1.9 million sentences, with 200 – 250 000
sentences per language. It covers a range of languages for which, to our knowledge, no other paraphrase dataset exists.

مجوز : Creative Commons Attribution 2.0 Generic
نسخه : 1.0.0
تقسیم ها :

شکاف	مثال ها
`'train'`	603

امکانات :

{
    "paraphrase_set_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "paraphrase": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lists": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "tags": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "language": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

ia

برای بارگذاری این مجموعه داده در TFDS از دستور زیر استفاده کنید:

ds = tfds.load('huggingface:tapaco/ia')

شرح :

A freely available paraphrase corpus for 73 languages extracted from the Tatoeba database. Tatoeba is a
crowdsourcing project mainly geared towards language learners. Its aim is to provide example sentences and translations for particular
linguistic constructions and words. The paraphrase corpus is created by populating a graph with Tatoeba sentences and equivalence links
between sentences “meaning the same thing”. This graph is then traversed to extract sets of paraphrases. Several language-independent
filters and pruning steps are applied to remove uninteresting sentences. A manual evaluation performed on three languages shows
that between half and three quarters of inferred paraphrases are correct and that most remaining ones are either correct but trivial, or
near-paraphrases that neutralize a morphological distinction. The corpus contains a total of 1.9 million sentences, with 200 – 250 000
sentences per language. It covers a range of languages for which, to our knowledge, no other paraphrase dataset exists.

مجوز : Creative Commons Attribution 2.0 Generic
نسخه : 1.0.0
تقسیم ها :

شکاف	مثال ها
`'train'`	2548

امکانات :

{
    "paraphrase_set_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "paraphrase": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lists": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "tags": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "language": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

شناسه

برای بارگذاری این مجموعه داده در TFDS از دستور زیر استفاده کنید:

ds = tfds.load('huggingface:tapaco/id')

شرح :

A freely available paraphrase corpus for 73 languages extracted from the Tatoeba database. Tatoeba is a
crowdsourcing project mainly geared towards language learners. Its aim is to provide example sentences and translations for particular
linguistic constructions and words. The paraphrase corpus is created by populating a graph with Tatoeba sentences and equivalence links
between sentences “meaning the same thing”. This graph is then traversed to extract sets of paraphrases. Several language-independent
filters and pruning steps are applied to remove uninteresting sentences. A manual evaluation performed on three languages shows
that between half and three quarters of inferred paraphrases are correct and that most remaining ones are either correct but trivial, or
near-paraphrases that neutralize a morphological distinction. The corpus contains a total of 1.9 million sentences, with 200 – 250 000
sentences per language. It covers a range of languages for which, to our knowledge, no other paraphrase dataset exists.

مجوز : Creative Commons Attribution 2.0 Generic
نسخه : 1.0.0
تقسیم ها :

شکاف	مثال ها
`'train'`	1602

امکانات :

{
    "paraphrase_set_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "paraphrase": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lists": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "tags": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "language": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

یعنی

برای بارگذاری این مجموعه داده در TFDS از دستور زیر استفاده کنید:

ds = tfds.load('huggingface:tapaco/ie')

شرح :

A freely available paraphrase corpus for 73 languages extracted from the Tatoeba database. Tatoeba is a
crowdsourcing project mainly geared towards language learners. Its aim is to provide example sentences and translations for particular
linguistic constructions and words. The paraphrase corpus is created by populating a graph with Tatoeba sentences and equivalence links
between sentences “meaning the same thing”. This graph is then traversed to extract sets of paraphrases. Several language-independent
filters and pruning steps are applied to remove uninteresting sentences. A manual evaluation performed on three languages shows
that between half and three quarters of inferred paraphrases are correct and that most remaining ones are either correct but trivial, or
near-paraphrases that neutralize a morphological distinction. The corpus contains a total of 1.9 million sentences, with 200 – 250 000
sentences per language. It covers a range of languages for which, to our knowledge, no other paraphrase dataset exists.

مجوز : Creative Commons Attribution 2.0 Generic
نسخه : 1.0.0
تقسیم ها :

شکاف	مثال ها
`'train'`	488

امکانات :

{
    "paraphrase_set_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "paraphrase": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lists": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "tags": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "language": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

io

برای بارگذاری این مجموعه داده در TFDS از دستور زیر استفاده کنید:

ds = tfds.load('huggingface:tapaco/io')

شرح :

A freely available paraphrase corpus for 73 languages extracted from the Tatoeba database. Tatoeba is a
crowdsourcing project mainly geared towards language learners. Its aim is to provide example sentences and translations for particular
linguistic constructions and words. The paraphrase corpus is created by populating a graph with Tatoeba sentences and equivalence links
between sentences “meaning the same thing”. This graph is then traversed to extract sets of paraphrases. Several language-independent
filters and pruning steps are applied to remove uninteresting sentences. A manual evaluation performed on three languages shows
that between half and three quarters of inferred paraphrases are correct and that most remaining ones are either correct but trivial, or
near-paraphrases that neutralize a morphological distinction. The corpus contains a total of 1.9 million sentences, with 200 – 250 000
sentences per language. It covers a range of languages for which, to our knowledge, no other paraphrase dataset exists.

مجوز : Creative Commons Attribution 2.0 Generic
نسخه : 1.0.0
تقسیم ها :

شکاف	مثال ها
`'train'`	480

امکانات :

{
    "paraphrase_set_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "paraphrase": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lists": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "tags": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "language": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

است

برای بارگذاری این مجموعه داده در TFDS از دستور زیر استفاده کنید:

ds = tfds.load('huggingface:tapaco/is')

شرح :

A freely available paraphrase corpus for 73 languages extracted from the Tatoeba database. Tatoeba is a
crowdsourcing project mainly geared towards language learners. Its aim is to provide example sentences and translations for particular
linguistic constructions and words. The paraphrase corpus is created by populating a graph with Tatoeba sentences and equivalence links
between sentences “meaning the same thing”. This graph is then traversed to extract sets of paraphrases. Several language-independent
filters and pruning steps are applied to remove uninteresting sentences. A manual evaluation performed on three languages shows
that between half and three quarters of inferred paraphrases are correct and that most remaining ones are either correct but trivial, or
near-paraphrases that neutralize a morphological distinction. The corpus contains a total of 1.9 million sentences, with 200 – 250 000
sentences per language. It covers a range of languages for which, to our knowledge, no other paraphrase dataset exists.

مجوز : Creative Commons Attribution 2.0 Generic
نسخه : 1.0.0
تقسیم ها :

شکاف	مثال ها
`'train'`	1641

امکانات :

{
    "paraphrase_set_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "paraphrase": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lists": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "tags": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "language": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

آی تی

برای بارگذاری این مجموعه داده در TFDS از دستور زیر استفاده کنید:

ds = tfds.load('huggingface:tapaco/it')

شرح :

A freely available paraphrase corpus for 73 languages extracted from the Tatoeba database. Tatoeba is a
crowdsourcing project mainly geared towards language learners. Its aim is to provide example sentences and translations for particular
linguistic constructions and words. The paraphrase corpus is created by populating a graph with Tatoeba sentences and equivalence links
between sentences “meaning the same thing”. This graph is then traversed to extract sets of paraphrases. Several language-independent
filters and pruning steps are applied to remove uninteresting sentences. A manual evaluation performed on three languages shows
that between half and three quarters of inferred paraphrases are correct and that most remaining ones are either correct but trivial, or
near-paraphrases that neutralize a morphological distinction. The corpus contains a total of 1.9 million sentences, with 200 – 250 000
sentences per language. It covers a range of languages for which, to our knowledge, no other paraphrase dataset exists.

مجوز : Creative Commons Attribution 2.0 Generic
نسخه : 1.0.0
تقسیم ها :

شکاف	مثال ها
`'train'`	198919

امکانات :

{
    "paraphrase_set_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "paraphrase": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lists": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "tags": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "language": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

ja

برای بارگذاری این مجموعه داده در TFDS از دستور زیر استفاده کنید:

ds = tfds.load('huggingface:tapaco/ja')

شرح :

A freely available paraphrase corpus for 73 languages extracted from the Tatoeba database. Tatoeba is a
crowdsourcing project mainly geared towards language learners. Its aim is to provide example sentences and translations for particular
linguistic constructions and words. The paraphrase corpus is created by populating a graph with Tatoeba sentences and equivalence links
between sentences “meaning the same thing”. This graph is then traversed to extract sets of paraphrases. Several language-independent
filters and pruning steps are applied to remove uninteresting sentences. A manual evaluation performed on three languages shows
that between half and three quarters of inferred paraphrases are correct and that most remaining ones are either correct but trivial, or
near-paraphrases that neutralize a morphological distinction. The corpus contains a total of 1.9 million sentences, with 200 – 250 000
sentences per language. It covers a range of languages for which, to our knowledge, no other paraphrase dataset exists.

مجوز : Creative Commons Attribution 2.0 Generic
نسخه : 1.0.0
تقسیم ها :

شکاف	مثال ها
`'train'`	44267

امکانات :

{
    "paraphrase_set_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "paraphrase": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lists": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "tags": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "language": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

jbo

برای بارگذاری این مجموعه داده در TFDS از دستور زیر استفاده کنید:

ds = tfds.load('huggingface:tapaco/jbo')

شرح :

A freely available paraphrase corpus for 73 languages extracted from the Tatoeba database. Tatoeba is a
crowdsourcing project mainly geared towards language learners. Its aim is to provide example sentences and translations for particular
linguistic constructions and words. The paraphrase corpus is created by populating a graph with Tatoeba sentences and equivalence links
between sentences “meaning the same thing”. This graph is then traversed to extract sets of paraphrases. Several language-independent
filters and pruning steps are applied to remove uninteresting sentences. A manual evaluation performed on three languages shows
that between half and three quarters of inferred paraphrases are correct and that most remaining ones are either correct but trivial, or
near-paraphrases that neutralize a morphological distinction. The corpus contains a total of 1.9 million sentences, with 200 – 250 000
sentences per language. It covers a range of languages for which, to our knowledge, no other paraphrase dataset exists.

مجوز : Creative Commons Attribution 2.0 Generic
نسخه : 1.0.0
تقسیم ها :

شکاف	مثال ها
`'train'`	2704

امکانات :

{
    "paraphrase_set_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "paraphrase": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lists": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "tags": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "language": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

کاب

برای بارگذاری این مجموعه داده در TFDS از دستور زیر استفاده کنید:

ds = tfds.load('huggingface:tapaco/kab')

شرح :

A freely available paraphrase corpus for 73 languages extracted from the Tatoeba database. Tatoeba is a
crowdsourcing project mainly geared towards language learners. Its aim is to provide example sentences and translations for particular
linguistic constructions and words. The paraphrase corpus is created by populating a graph with Tatoeba sentences and equivalence links
between sentences “meaning the same thing”. This graph is then traversed to extract sets of paraphrases. Several language-independent
filters and pruning steps are applied to remove uninteresting sentences. A manual evaluation performed on three languages shows
that between half and three quarters of inferred paraphrases are correct and that most remaining ones are either correct but trivial, or
near-paraphrases that neutralize a morphological distinction. The corpus contains a total of 1.9 million sentences, with 200 – 250 000
sentences per language. It covers a range of languages for which, to our knowledge, no other paraphrase dataset exists.

مجوز : Creative Commons Attribution 2.0 Generic
نسخه : 1.0.0
تقسیم ها :

شکاف	مثال ها
`'train'`	15944

امکانات :

{
    "paraphrase_set_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "paraphrase": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lists": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "tags": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "language": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

ko

برای بارگذاری این مجموعه داده در TFDS از دستور زیر استفاده کنید:

ds = tfds.load('huggingface:tapaco/ko')

شرح :

A freely available paraphrase corpus for 73 languages extracted from the Tatoeba database. Tatoeba is a
crowdsourcing project mainly geared towards language learners. Its aim is to provide example sentences and translations for particular
linguistic constructions and words. The paraphrase corpus is created by populating a graph with Tatoeba sentences and equivalence links
between sentences “meaning the same thing”. This graph is then traversed to extract sets of paraphrases. Several language-independent
filters and pruning steps are applied to remove uninteresting sentences. A manual evaluation performed on three languages shows
that between half and three quarters of inferred paraphrases are correct and that most remaining ones are either correct but trivial, or
near-paraphrases that neutralize a morphological distinction. The corpus contains a total of 1.9 million sentences, with 200 – 250 000
sentences per language. It covers a range of languages for which, to our knowledge, no other paraphrase dataset exists.

مجوز : Creative Commons Attribution 2.0 Generic
نسخه : 1.0.0
تقسیم ها :

شکاف	مثال ها
`'train'`	503

امکانات :

{
    "paraphrase_set_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "paraphrase": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lists": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "tags": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "language": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

کیلووات

برای بارگذاری این مجموعه داده در TFDS از دستور زیر استفاده کنید:

ds = tfds.load('huggingface:tapaco/kw')

شرح :

A freely available paraphrase corpus for 73 languages extracted from the Tatoeba database. Tatoeba is a
crowdsourcing project mainly geared towards language learners. Its aim is to provide example sentences and translations for particular
linguistic constructions and words. The paraphrase corpus is created by populating a graph with Tatoeba sentences and equivalence links
between sentences “meaning the same thing”. This graph is then traversed to extract sets of paraphrases. Several language-independent
filters and pruning steps are applied to remove uninteresting sentences. A manual evaluation performed on three languages shows
that between half and three quarters of inferred paraphrases are correct and that most remaining ones are either correct but trivial, or
near-paraphrases that neutralize a morphological distinction. The corpus contains a total of 1.9 million sentences, with 200 – 250 000
sentences per language. It covers a range of languages for which, to our knowledge, no other paraphrase dataset exists.

مجوز : Creative Commons Attribution 2.0 Generic
نسخه : 1.0.0
تقسیم ها :

شکاف	مثال ها
`'train'`	1328

امکانات :

{
    "paraphrase_set_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "paraphrase": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lists": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "tags": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "language": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

لا

برای بارگذاری این مجموعه داده در TFDS از دستور زیر استفاده کنید:

ds = tfds.load('huggingface:tapaco/la')

شرح :

A freely available paraphrase corpus for 73 languages extracted from the Tatoeba database. Tatoeba is a
crowdsourcing project mainly geared towards language learners. Its aim is to provide example sentences and translations for particular
linguistic constructions and words. The paraphrase corpus is created by populating a graph with Tatoeba sentences and equivalence links
between sentences “meaning the same thing”. This graph is then traversed to extract sets of paraphrases. Several language-independent
filters and pruning steps are applied to remove uninteresting sentences. A manual evaluation performed on three languages shows
that between half and three quarters of inferred paraphrases are correct and that most remaining ones are either correct but trivial, or
near-paraphrases that neutralize a morphological distinction. The corpus contains a total of 1.9 million sentences, with 200 – 250 000
sentences per language. It covers a range of languages for which, to our knowledge, no other paraphrase dataset exists.

مجوز : Creative Commons Attribution 2.0 Generic
نسخه : 1.0.0
تقسیم ها :

شکاف	مثال ها
`'train'`	6889

امکانات :

{
    "paraphrase_set_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "paraphrase": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lists": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "tags": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "language": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

lfn

برای بارگذاری این مجموعه داده در TFDS از دستور زیر استفاده کنید:

ds = tfds.load('huggingface:tapaco/lfn')

شرح :

A freely available paraphrase corpus for 73 languages extracted from the Tatoeba database. Tatoeba is a
crowdsourcing project mainly geared towards language learners. Its aim is to provide example sentences and translations for particular
linguistic constructions and words. The paraphrase corpus is created by populating a graph with Tatoeba sentences and equivalence links
between sentences “meaning the same thing”. This graph is then traversed to extract sets of paraphrases. Several language-independent
filters and pruning steps are applied to remove uninteresting sentences. A manual evaluation performed on three languages shows
that between half and three quarters of inferred paraphrases are correct and that most remaining ones are either correct but trivial, or
near-paraphrases that neutralize a morphological distinction. The corpus contains a total of 1.9 million sentences, with 200 – 250 000
sentences per language. It covers a range of languages for which, to our knowledge, no other paraphrase dataset exists.

مجوز : Creative Commons Attribution 2.0 Generic
نسخه : 1.0.0
تقسیم ها :

شکاف	مثال ها
`'train'`	2313

امکانات :

{
    "paraphrase_set_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "paraphrase": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lists": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "tags": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "language": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

آن

برای بارگذاری این مجموعه داده در TFDS از دستور زیر استفاده کنید:

ds = tfds.load('huggingface:tapaco/lt')

شرح :

A freely available paraphrase corpus for 73 languages extracted from the Tatoeba database. Tatoeba is a
crowdsourcing project mainly geared towards language learners. Its aim is to provide example sentences and translations for particular
linguistic constructions and words. The paraphrase corpus is created by populating a graph with Tatoeba sentences and equivalence links
between sentences “meaning the same thing”. This graph is then traversed to extract sets of paraphrases. Several language-independent
filters and pruning steps are applied to remove uninteresting sentences. A manual evaluation performed on three languages shows
that between half and three quarters of inferred paraphrases are correct and that most remaining ones are either correct but trivial, or
near-paraphrases that neutralize a morphological distinction. The corpus contains a total of 1.9 million sentences, with 200 – 250 000
sentences per language. It covers a range of languages for which, to our knowledge, no other paraphrase dataset exists.

مجوز : Creative Commons Attribution 2.0 Generic
نسخه : 1.0.0
تقسیم ها :

شکاف	مثال ها
`'train'`	8042

امکانات :

{
    "paraphrase_set_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "paraphrase": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lists": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "tags": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "language": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

mk

برای بارگذاری این مجموعه داده در TFDS از دستور زیر استفاده کنید:

ds = tfds.load('huggingface:tapaco/mk')

شرح :

A freely available paraphrase corpus for 73 languages extracted from the Tatoeba database. Tatoeba is a
crowdsourcing project mainly geared towards language learners. Its aim is to provide example sentences and translations for particular
linguistic constructions and words. The paraphrase corpus is created by populating a graph with Tatoeba sentences and equivalence links
between sentences “meaning the same thing”. This graph is then traversed to extract sets of paraphrases. Several language-independent
filters and pruning steps are applied to remove uninteresting sentences. A manual evaluation performed on three languages shows
that between half and three quarters of inferred paraphrases are correct and that most remaining ones are either correct but trivial, or
near-paraphrases that neutralize a morphological distinction. The corpus contains a total of 1.9 million sentences, with 200 – 250 000
sentences per language. It covers a range of languages for which, to our knowledge, no other paraphrase dataset exists.

مجوز : Creative Commons Attribution 2.0 Generic
نسخه : 1.0.0
تقسیم ها :

شکاف	مثال ها
`'train'`	14678

امکانات :

{
    "paraphrase_set_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "paraphrase": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lists": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "tags": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "language": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

آقای

برای بارگذاری این مجموعه داده در TFDS از دستور زیر استفاده کنید:

ds = tfds.load('huggingface:tapaco/mr')

شرح :

A freely available paraphrase corpus for 73 languages extracted from the Tatoeba database. Tatoeba is a
crowdsourcing project mainly geared towards language learners. Its aim is to provide example sentences and translations for particular
linguistic constructions and words. The paraphrase corpus is created by populating a graph with Tatoeba sentences and equivalence links
between sentences “meaning the same thing”. This graph is then traversed to extract sets of paraphrases. Several language-independent
filters and pruning steps are applied to remove uninteresting sentences. A manual evaluation performed on three languages shows
that between half and three quarters of inferred paraphrases are correct and that most remaining ones are either correct but trivial, or
near-paraphrases that neutralize a morphological distinction. The corpus contains a total of 1.9 million sentences, with 200 – 250 000
sentences per language. It covers a range of languages for which, to our knowledge, no other paraphrase dataset exists.

مجوز : Creative Commons Attribution 2.0 Generic
نسخه : 1.0.0
تقسیم ها :

شکاف	مثال ها
`'train'`	16413

امکانات :

{
    "paraphrase_set_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "paraphrase": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lists": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "tags": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "language": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

nb

برای بارگذاری این مجموعه داده در TFDS از دستور زیر استفاده کنید:

ds = tfds.load('huggingface:tapaco/nb')

شرح :

A freely available paraphrase corpus for 73 languages extracted from the Tatoeba database. Tatoeba is a
crowdsourcing project mainly geared towards language learners. Its aim is to provide example sentences and translations for particular
linguistic constructions and words. The paraphrase corpus is created by populating a graph with Tatoeba sentences and equivalence links
between sentences “meaning the same thing”. This graph is then traversed to extract sets of paraphrases. Several language-independent
filters and pruning steps are applied to remove uninteresting sentences. A manual evaluation performed on three languages shows
that between half and three quarters of inferred paraphrases are correct and that most remaining ones are either correct but trivial, or
near-paraphrases that neutralize a morphological distinction. The corpus contains a total of 1.9 million sentences, with 200 – 250 000
sentences per language. It covers a range of languages for which, to our knowledge, no other paraphrase dataset exists.

مجوز : Creative Commons Attribution 2.0 Generic
نسخه : 1.0.0
تقسیم ها :

شکاف	مثال ها
`'train'`	1094

امکانات :

{
    "paraphrase_set_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "paraphrase": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lists": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "tags": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "language": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

nds

برای بارگذاری این مجموعه داده در TFDS از دستور زیر استفاده کنید:

ds = tfds.load('huggingface:tapaco/nds')

شرح :

A freely available paraphrase corpus for 73 languages extracted from the Tatoeba database. Tatoeba is a
crowdsourcing project mainly geared towards language learners. Its aim is to provide example sentences and translations for particular
linguistic constructions and words. The paraphrase corpus is created by populating a graph with Tatoeba sentences and equivalence links
between sentences “meaning the same thing”. This graph is then traversed to extract sets of paraphrases. Several language-independent
filters and pruning steps are applied to remove uninteresting sentences. A manual evaluation performed on three languages shows
that between half and three quarters of inferred paraphrases are correct and that most remaining ones are either correct but trivial, or
near-paraphrases that neutralize a morphological distinction. The corpus contains a total of 1.9 million sentences, with 200 – 250 000
sentences per language. It covers a range of languages for which, to our knowledge, no other paraphrase dataset exists.

مجوز : Creative Commons Attribution 2.0 Generic
نسخه : 1.0.0
تقسیم ها :

شکاف	مثال ها
`'train'`	2633

امکانات :

{
    "paraphrase_set_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "paraphrase": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lists": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "tags": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "language": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

nl

برای بارگذاری این مجموعه داده در TFDS از دستور زیر استفاده کنید:

ds = tfds.load('huggingface:tapaco/nl')

شرح :

A freely available paraphrase corpus for 73 languages extracted from the Tatoeba database. Tatoeba is a
crowdsourcing project mainly geared towards language learners. Its aim is to provide example sentences and translations for particular
linguistic constructions and words. The paraphrase corpus is created by populating a graph with Tatoeba sentences and equivalence links
between sentences “meaning the same thing”. This graph is then traversed to extract sets of paraphrases. Several language-independent
filters and pruning steps are applied to remove uninteresting sentences. A manual evaluation performed on three languages shows
that between half and three quarters of inferred paraphrases are correct and that most remaining ones are either correct but trivial, or
near-paraphrases that neutralize a morphological distinction. The corpus contains a total of 1.9 million sentences, with 200 – 250 000
sentences per language. It covers a range of languages for which, to our knowledge, no other paraphrase dataset exists.

مجوز : Creative Commons Attribution 2.0 Generic
نسخه : 1.0.0
تقسیم ها :

شکاف	مثال ها
`'train'`	23561

امکانات :

{
    "paraphrase_set_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "paraphrase": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lists": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "tags": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "language": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

orv

برای بارگذاری این مجموعه داده در TFDS از دستور زیر استفاده کنید:

ds = tfds.load('huggingface:tapaco/orv')

شرح :

A freely available paraphrase corpus for 73 languages extracted from the Tatoeba database. Tatoeba is a
crowdsourcing project mainly geared towards language learners. Its aim is to provide example sentences and translations for particular
linguistic constructions and words. The paraphrase corpus is created by populating a graph with Tatoeba sentences and equivalence links
between sentences “meaning the same thing”. This graph is then traversed to extract sets of paraphrases. Several language-independent
filters and pruning steps are applied to remove uninteresting sentences. A manual evaluation performed on three languages shows
that between half and three quarters of inferred paraphrases are correct and that most remaining ones are either correct but trivial, or
near-paraphrases that neutralize a morphological distinction. The corpus contains a total of 1.9 million sentences, with 200 – 250 000
sentences per language. It covers a range of languages for which, to our knowledge, no other paraphrase dataset exists.

مجوز : Creative Commons Attribution 2.0 Generic
نسخه : 1.0.0
تقسیم ها :

شکاف	مثال ها
`'train'`	471

امکانات :

{
    "paraphrase_set_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "paraphrase": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lists": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "tags": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "language": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

اوتا

برای بارگذاری این مجموعه داده در TFDS از دستور زیر استفاده کنید:

ds = tfds.load('huggingface:tapaco/ota')

شرح :

A freely available paraphrase corpus for 73 languages extracted from the Tatoeba database. Tatoeba is a
crowdsourcing project mainly geared towards language learners. Its aim is to provide example sentences and translations for particular
linguistic constructions and words. The paraphrase corpus is created by populating a graph with Tatoeba sentences and equivalence links
between sentences “meaning the same thing”. This graph is then traversed to extract sets of paraphrases. Several language-independent
filters and pruning steps are applied to remove uninteresting sentences. A manual evaluation performed on three languages shows
that between half and three quarters of inferred paraphrases are correct and that most remaining ones are either correct but trivial, or
near-paraphrases that neutralize a morphological distinction. The corpus contains a total of 1.9 million sentences, with 200 – 250 000
sentences per language. It covers a range of languages for which, to our knowledge, no other paraphrase dataset exists.

مجوز : Creative Commons Attribution 2.0 Generic
نسخه : 1.0.0
تقسیم ها :

شکاف	مثال ها
`'train'`	486

امکانات :

{
    "paraphrase_set_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "paraphrase": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lists": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "tags": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "language": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

pes

برای بارگذاری این مجموعه داده در TFDS از دستور زیر استفاده کنید:

ds = tfds.load('huggingface:tapaco/pes')

شرح :

A freely available paraphrase corpus for 73 languages extracted from the Tatoeba database. Tatoeba is a
crowdsourcing project mainly geared towards language learners. Its aim is to provide example sentences and translations for particular
linguistic constructions and words. The paraphrase corpus is created by populating a graph with Tatoeba sentences and equivalence links
between sentences “meaning the same thing”. This graph is then traversed to extract sets of paraphrases. Several language-independent
filters and pruning steps are applied to remove uninteresting sentences. A manual evaluation performed on three languages shows
that between half and three quarters of inferred paraphrases are correct and that most remaining ones are either correct but trivial, or
near-paraphrases that neutralize a morphological distinction. The corpus contains a total of 1.9 million sentences, with 200 – 250 000
sentences per language. It covers a range of languages for which, to our knowledge, no other paraphrase dataset exists.

مجوز : Creative Commons Attribution 2.0 Generic
نسخه : 1.0.0
تقسیم ها :

شکاف	مثال ها
`'train'`	4285

امکانات :

{
    "paraphrase_set_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "paraphrase": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lists": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "tags": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "language": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

pl

برای بارگذاری این مجموعه داده در TFDS از دستور زیر استفاده کنید:

ds = tfds.load('huggingface:tapaco/pl')

شرح :

A freely available paraphrase corpus for 73 languages extracted from the Tatoeba database. Tatoeba is a
crowdsourcing project mainly geared towards language learners. Its aim is to provide example sentences and translations for particular
linguistic constructions and words. The paraphrase corpus is created by populating a graph with Tatoeba sentences and equivalence links
between sentences “meaning the same thing”. This graph is then traversed to extract sets of paraphrases. Several language-independent
filters and pruning steps are applied to remove uninteresting sentences. A manual evaluation performed on three languages shows
that between half and three quarters of inferred paraphrases are correct and that most remaining ones are either correct but trivial, or
near-paraphrases that neutralize a morphological distinction. The corpus contains a total of 1.9 million sentences, with 200 – 250 000
sentences per language. It covers a range of languages for which, to our knowledge, no other paraphrase dataset exists.

مجوز : Creative Commons Attribution 2.0 Generic
نسخه : 1.0.0
تقسیم ها :

شکاف	مثال ها
`'train'`	22391

امکانات :

{
    "paraphrase_set_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "paraphrase": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lists": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "tags": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "language": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

pt

برای بارگذاری این مجموعه داده در TFDS از دستور زیر استفاده کنید:

ds = tfds.load('huggingface:tapaco/pt')

شرح :

A freely available paraphrase corpus for 73 languages extracted from the Tatoeba database. Tatoeba is a
crowdsourcing project mainly geared towards language learners. Its aim is to provide example sentences and translations for particular
linguistic constructions and words. The paraphrase corpus is created by populating a graph with Tatoeba sentences and equivalence links
between sentences “meaning the same thing”. This graph is then traversed to extract sets of paraphrases. Several language-independent
filters and pruning steps are applied to remove uninteresting sentences. A manual evaluation performed on three languages shows
that between half and three quarters of inferred paraphrases are correct and that most remaining ones are either correct but trivial, or
near-paraphrases that neutralize a morphological distinction. The corpus contains a total of 1.9 million sentences, with 200 – 250 000
sentences per language. It covers a range of languages for which, to our knowledge, no other paraphrase dataset exists.

مجوز : Creative Commons Attribution 2.0 Generic
نسخه : 1.0.0
تقسیم ها :

شکاف	مثال ها
`'train'`	78430

امکانات :

{
    "paraphrase_set_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "paraphrase": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lists": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "tags": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "language": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

rn

برای بارگذاری این مجموعه داده در TFDS از دستور زیر استفاده کنید:

ds = tfds.load('huggingface:tapaco/rn')

شرح :

A freely available paraphrase corpus for 73 languages extracted from the Tatoeba database. Tatoeba is a
crowdsourcing project mainly geared towards language learners. Its aim is to provide example sentences and translations for particular
linguistic constructions and words. The paraphrase corpus is created by populating a graph with Tatoeba sentences and equivalence links
between sentences “meaning the same thing”. This graph is then traversed to extract sets of paraphrases. Several language-independent
filters and pruning steps are applied to remove uninteresting sentences. A manual evaluation performed on three languages shows
that between half and three quarters of inferred paraphrases are correct and that most remaining ones are either correct but trivial, or
near-paraphrases that neutralize a morphological distinction. The corpus contains a total of 1.9 million sentences, with 200 – 250 000
sentences per language. It covers a range of languages for which, to our knowledge, no other paraphrase dataset exists.

مجوز : Creative Commons Attribution 2.0 Generic
نسخه : 1.0.0
تقسیم ها :

شکاف	مثال ها
`'train'`	648

امکانات :

{
    "paraphrase_set_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "paraphrase": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lists": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "tags": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "language": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

ro

برای بارگذاری این مجموعه داده در TFDS از دستور زیر استفاده کنید:

ds = tfds.load('huggingface:tapaco/ro')

شرح :

A freely available paraphrase corpus for 73 languages extracted from the Tatoeba database. Tatoeba is a
crowdsourcing project mainly geared towards language learners. Its aim is to provide example sentences and translations for particular
linguistic constructions and words. The paraphrase corpus is created by populating a graph with Tatoeba sentences and equivalence links
between sentences “meaning the same thing”. This graph is then traversed to extract sets of paraphrases. Several language-independent
filters and pruning steps are applied to remove uninteresting sentences. A manual evaluation performed on three languages shows
that between half and three quarters of inferred paraphrases are correct and that most remaining ones are either correct but trivial, or
near-paraphrases that neutralize a morphological distinction. The corpus contains a total of 1.9 million sentences, with 200 – 250 000
sentences per language. It covers a range of languages for which, to our knowledge, no other paraphrase dataset exists.

مجوز : Creative Commons Attribution 2.0 Generic
نسخه : 1.0.0
تقسیم ها :

شکاف	مثال ها
`'train'`	2092

امکانات :

{
    "paraphrase_set_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "paraphrase": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lists": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "tags": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "language": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

ru

برای بارگذاری این مجموعه داده در TFDS از دستور زیر استفاده کنید:

ds = tfds.load('huggingface:tapaco/ru')

شرح :

A freely available paraphrase corpus for 73 languages extracted from the Tatoeba database. Tatoeba is a
crowdsourcing project mainly geared towards language learners. Its aim is to provide example sentences and translations for particular
linguistic constructions and words. The paraphrase corpus is created by populating a graph with Tatoeba sentences and equivalence links
between sentences “meaning the same thing”. This graph is then traversed to extract sets of paraphrases. Several language-independent
filters and pruning steps are applied to remove uninteresting sentences. A manual evaluation performed on three languages shows
that between half and three quarters of inferred paraphrases are correct and that most remaining ones are either correct but trivial, or
near-paraphrases that neutralize a morphological distinction. The corpus contains a total of 1.9 million sentences, with 200 – 250 000
sentences per language. It covers a range of languages for which, to our knowledge, no other paraphrase dataset exists.

مجوز : Creative Commons Attribution 2.0 Generic
نسخه : 1.0.0
تقسیم ها :

شکاف	مثال ها
`'train'`	251263

امکانات :

{
    "paraphrase_set_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "paraphrase": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lists": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "tags": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "language": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

sl

برای بارگذاری این مجموعه داده در TFDS از دستور زیر استفاده کنید:

ds = tfds.load('huggingface:tapaco/sl')

شرح :

A freely available paraphrase corpus for 73 languages extracted from the Tatoeba database. Tatoeba is a
crowdsourcing project mainly geared towards language learners. Its aim is to provide example sentences and translations for particular
linguistic constructions and words. The paraphrase corpus is created by populating a graph with Tatoeba sentences and equivalence links
between sentences “meaning the same thing”. This graph is then traversed to extract sets of paraphrases. Several language-independent
filters and pruning steps are applied to remove uninteresting sentences. A manual evaluation performed on three languages shows
that between half and three quarters of inferred paraphrases are correct and that most remaining ones are either correct but trivial, or
near-paraphrases that neutralize a morphological distinction. The corpus contains a total of 1.9 million sentences, with 200 – 250 000
sentences per language. It covers a range of languages for which, to our knowledge, no other paraphrase dataset exists.

مجوز : Creative Commons Attribution 2.0 Generic
نسخه : 1.0.0
تقسیم ها :

شکاف	مثال ها
`'train'`	706

امکانات :

{
    "paraphrase_set_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "paraphrase": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lists": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "tags": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "language": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

sr

برای بارگذاری این مجموعه داده در TFDS از دستور زیر استفاده کنید:

ds = tfds.load('huggingface:tapaco/sr')

شرح :

A freely available paraphrase corpus for 73 languages extracted from the Tatoeba database. Tatoeba is a
crowdsourcing project mainly geared towards language learners. Its aim is to provide example sentences and translations for particular
linguistic constructions and words. The paraphrase corpus is created by populating a graph with Tatoeba sentences and equivalence links
between sentences “meaning the same thing”. This graph is then traversed to extract sets of paraphrases. Several language-independent
filters and pruning steps are applied to remove uninteresting sentences. A manual evaluation performed on three languages shows
that between half and three quarters of inferred paraphrases are correct and that most remaining ones are either correct but trivial, or
near-paraphrases that neutralize a morphological distinction. The corpus contains a total of 1.9 million sentences, with 200 – 250 000
sentences per language. It covers a range of languages for which, to our knowledge, no other paraphrase dataset exists.

مجوز : Creative Commons Attribution 2.0 Generic
نسخه : 1.0.0
تقسیم ها :

شکاف	مثال ها
`'train'`	8175

امکانات :

{
    "paraphrase_set_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "paraphrase": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lists": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "tags": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "language": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

sv

برای بارگذاری این مجموعه داده در TFDS از دستور زیر استفاده کنید:

ds = tfds.load('huggingface:tapaco/sv')

شرح :

A freely available paraphrase corpus for 73 languages extracted from the Tatoeba database. Tatoeba is a
crowdsourcing project mainly geared towards language learners. Its aim is to provide example sentences and translations for particular
linguistic constructions and words. The paraphrase corpus is created by populating a graph with Tatoeba sentences and equivalence links
between sentences “meaning the same thing”. This graph is then traversed to extract sets of paraphrases. Several language-independent
filters and pruning steps are applied to remove uninteresting sentences. A manual evaluation performed on three languages shows
that between half and three quarters of inferred paraphrases are correct and that most remaining ones are either correct but trivial, or
near-paraphrases that neutralize a morphological distinction. The corpus contains a total of 1.9 million sentences, with 200 – 250 000
sentences per language. It covers a range of languages for which, to our knowledge, no other paraphrase dataset exists.

مجوز : Creative Commons Attribution 2.0 Generic
نسخه : 1.0.0
تقسیم ها :

شکاف	مثال ها
`'train'`	7005

امکانات :

{
    "paraphrase_set_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "paraphrase": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lists": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "tags": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "language": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

tk

برای بارگذاری این مجموعه داده در TFDS از دستور زیر استفاده کنید:

ds = tfds.load('huggingface:tapaco/tk')

شرح :

A freely available paraphrase corpus for 73 languages extracted from the Tatoeba database. Tatoeba is a
crowdsourcing project mainly geared towards language learners. Its aim is to provide example sentences and translations for particular
linguistic constructions and words. The paraphrase corpus is created by populating a graph with Tatoeba sentences and equivalence links
between sentences “meaning the same thing”. This graph is then traversed to extract sets of paraphrases. Several language-independent
filters and pruning steps are applied to remove uninteresting sentences. A manual evaluation performed on three languages shows
that between half and three quarters of inferred paraphrases are correct and that most remaining ones are either correct but trivial, or
near-paraphrases that neutralize a morphological distinction. The corpus contains a total of 1.9 million sentences, with 200 – 250 000
sentences per language. It covers a range of languages for which, to our knowledge, no other paraphrase dataset exists.

مجوز : Creative Commons Attribution 2.0 Generic
نسخه : 1.0.0
تقسیم ها :

شکاف	مثال ها
`'train'`	1165

امکانات :

{
    "paraphrase_set_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "paraphrase": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lists": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "tags": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "language": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

tl

برای بارگذاری این مجموعه داده در TFDS از دستور زیر استفاده کنید:

ds = tfds.load('huggingface:tapaco/tl')

شرح :

A freely available paraphrase corpus for 73 languages extracted from the Tatoeba database. Tatoeba is a
crowdsourcing project mainly geared towards language learners. Its aim is to provide example sentences and translations for particular
linguistic constructions and words. The paraphrase corpus is created by populating a graph with Tatoeba sentences and equivalence links
between sentences “meaning the same thing”. This graph is then traversed to extract sets of paraphrases. Several language-independent
filters and pruning steps are applied to remove uninteresting sentences. A manual evaluation performed on three languages shows
that between half and three quarters of inferred paraphrases are correct and that most remaining ones are either correct but trivial, or
near-paraphrases that neutralize a morphological distinction. The corpus contains a total of 1.9 million sentences, with 200 – 250 000
sentences per language. It covers a range of languages for which, to our knowledge, no other paraphrase dataset exists.

مجوز : Creative Commons Attribution 2.0 Generic
نسخه : 1.0.0
تقسیم ها :

شکاف	مثال ها
`'train'`	1017

امکانات :

{
    "paraphrase_set_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "paraphrase": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lists": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "tags": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "language": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

tlh

برای بارگذاری این مجموعه داده در TFDS از دستور زیر استفاده کنید:

ds = tfds.load('huggingface:tapaco/tlh')

شرح :

A freely available paraphrase corpus for 73 languages extracted from the Tatoeba database. Tatoeba is a
crowdsourcing project mainly geared towards language learners. Its aim is to provide example sentences and translations for particular
linguistic constructions and words. The paraphrase corpus is created by populating a graph with Tatoeba sentences and equivalence links
between sentences “meaning the same thing”. This graph is then traversed to extract sets of paraphrases. Several language-independent
filters and pruning steps are applied to remove uninteresting sentences. A manual evaluation performed on three languages shows
that between half and three quarters of inferred paraphrases are correct and that most remaining ones are either correct but trivial, or
near-paraphrases that neutralize a morphological distinction. The corpus contains a total of 1.9 million sentences, with 200 – 250 000
sentences per language. It covers a range of languages for which, to our knowledge, no other paraphrase dataset exists.

مجوز : Creative Commons Attribution 2.0 Generic
نسخه : 1.0.0
تقسیم ها :

شکاف	مثال ها
`'train'`	2804

امکانات :

{
    "paraphrase_set_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "paraphrase": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lists": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "tags": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "language": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

توکی

برای بارگذاری این مجموعه داده در TFDS از دستور زیر استفاده کنید:

ds = tfds.load('huggingface:tapaco/toki')

شرح :

A freely available paraphrase corpus for 73 languages extracted from the Tatoeba database. Tatoeba is a
crowdsourcing project mainly geared towards language learners. Its aim is to provide example sentences and translations for particular
linguistic constructions and words. The paraphrase corpus is created by populating a graph with Tatoeba sentences and equivalence links
between sentences “meaning the same thing”. This graph is then traversed to extract sets of paraphrases. Several language-independent
filters and pruning steps are applied to remove uninteresting sentences. A manual evaluation performed on three languages shows
that between half and three quarters of inferred paraphrases are correct and that most remaining ones are either correct but trivial, or
near-paraphrases that neutralize a morphological distinction. The corpus contains a total of 1.9 million sentences, with 200 – 250 000
sentences per language. It covers a range of languages for which, to our knowledge, no other paraphrase dataset exists.

مجوز : Creative Commons Attribution 2.0 Generic
نسخه : 1.0.0
تقسیم ها :

شکاف	مثال ها
`'train'`	3738

امکانات :

{
    "paraphrase_set_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "paraphrase": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lists": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "tags": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "language": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

tr

برای بارگذاری این مجموعه داده در TFDS از دستور زیر استفاده کنید:

ds = tfds.load('huggingface:tapaco/tr')

شرح :

A freely available paraphrase corpus for 73 languages extracted from the Tatoeba database. Tatoeba is a
crowdsourcing project mainly geared towards language learners. Its aim is to provide example sentences and translations for particular
linguistic constructions and words. The paraphrase corpus is created by populating a graph with Tatoeba sentences and equivalence links
between sentences “meaning the same thing”. This graph is then traversed to extract sets of paraphrases. Several language-independent
filters and pruning steps are applied to remove uninteresting sentences. A manual evaluation performed on three languages shows
that between half and three quarters of inferred paraphrases are correct and that most remaining ones are either correct but trivial, or
near-paraphrases that neutralize a morphological distinction. The corpus contains a total of 1.9 million sentences, with 200 – 250 000
sentences per language. It covers a range of languages for which, to our knowledge, no other paraphrase dataset exists.

مجوز : Creative Commons Attribution 2.0 Generic
نسخه : 1.0.0
تقسیم ها :

شکاف	مثال ها
`'train'`	142088

امکانات :

{
    "paraphrase_set_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "paraphrase": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lists": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "tags": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "language": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

tt

برای بارگذاری این مجموعه داده در TFDS از دستور زیر استفاده کنید:

ds = tfds.load('huggingface:tapaco/tt')

شرح :

A freely available paraphrase corpus for 73 languages extracted from the Tatoeba database. Tatoeba is a
crowdsourcing project mainly geared towards language learners. Its aim is to provide example sentences and translations for particular
linguistic constructions and words. The paraphrase corpus is created by populating a graph with Tatoeba sentences and equivalence links
between sentences “meaning the same thing”. This graph is then traversed to extract sets of paraphrases. Several language-independent
filters and pruning steps are applied to remove uninteresting sentences. A manual evaluation performed on three languages shows
that between half and three quarters of inferred paraphrases are correct and that most remaining ones are either correct but trivial, or
near-paraphrases that neutralize a morphological distinction. The corpus contains a total of 1.9 million sentences, with 200 – 250 000
sentences per language. It covers a range of languages for which, to our knowledge, no other paraphrase dataset exists.

مجوز : Creative Commons Attribution 2.0 Generic
نسخه : 1.0.0
تقسیم ها :

شکاف	مثال ها
`'train'`	2398

امکانات :

{
    "paraphrase_set_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "paraphrase": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lists": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "tags": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "language": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

ug

برای بارگذاری این مجموعه داده در TFDS از دستور زیر استفاده کنید:

ds = tfds.load('huggingface:tapaco/ug')

شرح :

A freely available paraphrase corpus for 73 languages extracted from the Tatoeba database. Tatoeba is a
crowdsourcing project mainly geared towards language learners. Its aim is to provide example sentences and translations for particular
linguistic constructions and words. The paraphrase corpus is created by populating a graph with Tatoeba sentences and equivalence links
between sentences “meaning the same thing”. This graph is then traversed to extract sets of paraphrases. Several language-independent
filters and pruning steps are applied to remove uninteresting sentences. A manual evaluation performed on three languages shows
that between half and three quarters of inferred paraphrases are correct and that most remaining ones are either correct but trivial, or
near-paraphrases that neutralize a morphological distinction. The corpus contains a total of 1.9 million sentences, with 200 – 250 000
sentences per language. It covers a range of languages for which, to our knowledge, no other paraphrase dataset exists.

مجوز : Creative Commons Attribution 2.0 Generic
نسخه : 1.0.0
تقسیم ها :

شکاف	مثال ها
`'train'`	1183

امکانات :

{
    "paraphrase_set_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "paraphrase": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lists": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "tags": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "language": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

انگلستان

برای بارگذاری این مجموعه داده در TFDS از دستور زیر استفاده کنید:

ds = tfds.load('huggingface:tapaco/uk')

شرح :

A freely available paraphrase corpus for 73 languages extracted from the Tatoeba database. Tatoeba is a
crowdsourcing project mainly geared towards language learners. Its aim is to provide example sentences and translations for particular
linguistic constructions and words. The paraphrase corpus is created by populating a graph with Tatoeba sentences and equivalence links
between sentences “meaning the same thing”. This graph is then traversed to extract sets of paraphrases. Several language-independent
filters and pruning steps are applied to remove uninteresting sentences. A manual evaluation performed on three languages shows
that between half and three quarters of inferred paraphrases are correct and that most remaining ones are either correct but trivial, or
near-paraphrases that neutralize a morphological distinction. The corpus contains a total of 1.9 million sentences, with 200 – 250 000
sentences per language. It covers a range of languages for which, to our knowledge, no other paraphrase dataset exists.

مجوز : Creative Commons Attribution 2.0 Generic
نسخه : 1.0.0
تقسیم ها :

شکاف	مثال ها
`'train'`	54431

امکانات :

{
    "paraphrase_set_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "paraphrase": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lists": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "tags": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "language": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

ur

برای بارگذاری این مجموعه داده در TFDS از دستور زیر استفاده کنید:

ds = tfds.load('huggingface:tapaco/ur')

شرح :

A freely available paraphrase corpus for 73 languages extracted from the Tatoeba database. Tatoeba is a
crowdsourcing project mainly geared towards language learners. Its aim is to provide example sentences and translations for particular
linguistic constructions and words. The paraphrase corpus is created by populating a graph with Tatoeba sentences and equivalence links
between sentences “meaning the same thing”. This graph is then traversed to extract sets of paraphrases. Several language-independent
filters and pruning steps are applied to remove uninteresting sentences. A manual evaluation performed on three languages shows
that between half and three quarters of inferred paraphrases are correct and that most remaining ones are either correct but trivial, or
near-paraphrases that neutralize a morphological distinction. The corpus contains a total of 1.9 million sentences, with 200 – 250 000
sentences per language. It covers a range of languages for which, to our knowledge, no other paraphrase dataset exists.

مجوز : Creative Commons Attribution 2.0 Generic
نسخه : 1.0.0
تقسیم ها :

شکاف	مثال ها
`'train'`	252

امکانات :

{
    "paraphrase_set_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "paraphrase": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lists": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "tags": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "language": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

vi

برای بارگذاری این مجموعه داده در TFDS از دستور زیر استفاده کنید:

ds = tfds.load('huggingface:tapaco/vi')

شرح :

A freely available paraphrase corpus for 73 languages extracted from the Tatoeba database. Tatoeba is a
crowdsourcing project mainly geared towards language learners. Its aim is to provide example sentences and translations for particular
linguistic constructions and words. The paraphrase corpus is created by populating a graph with Tatoeba sentences and equivalence links
between sentences “meaning the same thing”. This graph is then traversed to extract sets of paraphrases. Several language-independent
filters and pruning steps are applied to remove uninteresting sentences. A manual evaluation performed on three languages shows
that between half and three quarters of inferred paraphrases are correct and that most remaining ones are either correct but trivial, or
near-paraphrases that neutralize a morphological distinction. The corpus contains a total of 1.9 million sentences, with 200 – 250 000
sentences per language. It covers a range of languages for which, to our knowledge, no other paraphrase dataset exists.

مجوز : Creative Commons Attribution 2.0 Generic
نسخه : 1.0.0
تقسیم ها :

شکاف	مثال ها
`'train'`	962

امکانات :

{
    "paraphrase_set_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "paraphrase": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lists": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "tags": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "language": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

vo

برای بارگذاری این مجموعه داده در TFDS از دستور زیر استفاده کنید:

ds = tfds.load('huggingface:tapaco/vo')

شرح :

A freely available paraphrase corpus for 73 languages extracted from the Tatoeba database. Tatoeba is a
crowdsourcing project mainly geared towards language learners. Its aim is to provide example sentences and translations for particular
linguistic constructions and words. The paraphrase corpus is created by populating a graph with Tatoeba sentences and equivalence links
between sentences “meaning the same thing”. This graph is then traversed to extract sets of paraphrases. Several language-independent
filters and pruning steps are applied to remove uninteresting sentences. A manual evaluation performed on three languages shows
that between half and three quarters of inferred paraphrases are correct and that most remaining ones are either correct but trivial, or
near-paraphrases that neutralize a morphological distinction. The corpus contains a total of 1.9 million sentences, with 200 – 250 000
sentences per language. It covers a range of languages for which, to our knowledge, no other paraphrase dataset exists.

مجوز : Creative Commons Attribution 2.0 Generic
نسخه : 1.0.0
تقسیم ها :

شکاف	مثال ها
`'train'`	328

امکانات :

{
    "paraphrase_set_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "paraphrase": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lists": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "tags": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "language": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

جنگ

برای بارگذاری این مجموعه داده در TFDS از دستور زیر استفاده کنید:

ds = tfds.load('huggingface:tapaco/war')

شرح :

A freely available paraphrase corpus for 73 languages extracted from the Tatoeba database. Tatoeba is a
crowdsourcing project mainly geared towards language learners. Its aim is to provide example sentences and translations for particular
linguistic constructions and words. The paraphrase corpus is created by populating a graph with Tatoeba sentences and equivalence links
between sentences “meaning the same thing”. This graph is then traversed to extract sets of paraphrases. Several language-independent
filters and pruning steps are applied to remove uninteresting sentences. A manual evaluation performed on three languages shows
that between half and three quarters of inferred paraphrases are correct and that most remaining ones are either correct but trivial, or
near-paraphrases that neutralize a morphological distinction. The corpus contains a total of 1.9 million sentences, with 200 – 250 000
sentences per language. It covers a range of languages for which, to our knowledge, no other paraphrase dataset exists.

مجوز : Creative Commons Attribution 2.0 Generic
نسخه : 1.0.0
تقسیم ها :

شکاف	مثال ها
`'train'`	327

امکانات :

{
    "paraphrase_set_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "paraphrase": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lists": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "tags": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "language": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

وو

برای بارگذاری این مجموعه داده در TFDS از دستور زیر استفاده کنید:

ds = tfds.load('huggingface:tapaco/wuu')

شرح :

A freely available paraphrase corpus for 73 languages extracted from the Tatoeba database. Tatoeba is a
crowdsourcing project mainly geared towards language learners. Its aim is to provide example sentences and translations for particular
linguistic constructions and words. The paraphrase corpus is created by populating a graph with Tatoeba sentences and equivalence links
between sentences “meaning the same thing”. This graph is then traversed to extract sets of paraphrases. Several language-independent
filters and pruning steps are applied to remove uninteresting sentences. A manual evaluation performed on three languages shows
that between half and three quarters of inferred paraphrases are correct and that most remaining ones are either correct but trivial, or
near-paraphrases that neutralize a morphological distinction. The corpus contains a total of 1.9 million sentences, with 200 – 250 000
sentences per language. It covers a range of languages for which, to our knowledge, no other paraphrase dataset exists.

مجوز : Creative Commons Attribution 2.0 Generic
نسخه : 1.0.0
تقسیم ها :

شکاف	مثال ها
`'train'`	408

امکانات :

{
    "paraphrase_set_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "paraphrase": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lists": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "tags": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "language": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

بله

برای بارگذاری این مجموعه داده در TFDS از دستور زیر استفاده کنید:

ds = tfds.load('huggingface:tapaco/yue')

شرح :

A freely available paraphrase corpus for 73 languages extracted from the Tatoeba database. Tatoeba is a
crowdsourcing project mainly geared towards language learners. Its aim is to provide example sentences and translations for particular
linguistic constructions and words. The paraphrase corpus is created by populating a graph with Tatoeba sentences and equivalence links
between sentences “meaning the same thing”. This graph is then traversed to extract sets of paraphrases. Several language-independent
filters and pruning steps are applied to remove uninteresting sentences. A manual evaluation performed on three languages shows
that between half and three quarters of inferred paraphrases are correct and that most remaining ones are either correct but trivial, or
near-paraphrases that neutralize a morphological distinction. The corpus contains a total of 1.9 million sentences, with 200 – 250 000
sentences per language. It covers a range of languages for which, to our knowledge, no other paraphrase dataset exists.

مجوز : Creative Commons Attribution 2.0 Generic
نسخه : 1.0.0
تقسیم ها :

شکاف	مثال ها
`'train'`	561

امکانات :

{
    "paraphrase_set_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "paraphrase": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lists": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "tags": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "language": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}